Structures of Pseudomonas aeruginosa LpxA Reveal the Basis for Its

Subscriber access provided by Stockholm University Library

Article

Structures of Pseudomonas aeruginosa LpxA reveal basis for substrate selectivity Emmanuel W. Smith, Xiujun Zhang, Cyrus Behzadi, Logan D Andrews, Frederick Cohen, and Yu Chen Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.5b00720 • Publication Date (Web): 09 Sep 2015 Downloaded from http://pubs.acs.org on September 11, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Structures of Pseudomonas aeruginosa LpxA reveal basis for substrate selectivity Funding Source Statement This work is funded by a collaboration grant from Achaogen Inc.

Emmanuel W. Smith,1 XiuJun Zhang,1 Cyrus Behzadi,1 Logan D. Andrews,2 Frederick Cohen,2 and Yu Chen*1 1

Department of Molecular Medicine, University of South Florida, 12901 Bruce B. Downs

Blvd., Tampa, FL 33612, USA 2

ACHAOGEN Inc., 7000 Shoreline Court, South San Francisco, CA 94080, USA

*

Author to whom correspondence should be addressed: Yu Chen at (813) 974-7809; Fax (813) 974-7357; E-mail: [email protected]

1 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 45

Abbreviations ACP acyl carrier protein LβH left-handed parallel beta-helix LPS lipopolysaccharide LpxA UDP-N-acetylglucosamine acyltransferase RMSD root-mean-square-deviation UDP-GlcNAc uridine diphosphate N-acetylglucosamine


Page 3 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

ABSTRACT

In Gram-negative bacteria, the first step of lipid A biosynthesis is catalyzed by UDP-Nacetylglucosamine acyltransferase (LpxA) through the transfer of a R-3-hydroxyacyl chain from the acyl carrier protein (ACP) to the 3-hydroxyl group of UDP-GlcNAc. Previous studies suggest that LpxA is a critical determinant of the acyl chain length found in lipid A, which varies between species of bacteria. In E. coli and L. interrogans, LpxA prefers to incorporate longer R-3-hydroxyacyl chains (C14 and C12 respectively), while in P. aeruginosa the enzyme is selective for R-3-hydroxydecanoyl, a 10-hydrocarbon long acyl chain. We now report three P. aeruginosa LpxA crystal structures; apo protein, substrate complex with UDP-GlcNAc, and product complex with UDP-3-O-(R-3hydroxydecanoyl)-GlcNAc.

A comparison between the apo form and complexes

identifies key residues that position UDP-GlcNAc appropriately for catalysis, and supports the role of the catalytic His121 in activating the UDP-GlcNAc 3-hydroxyl group for nucleophilic attack during the reaction. The product-complex structure for the first time offers structural insights into how Met169 serves to constrain the length of the acyl chain and thus functions as the so-called ‘hydrocarbon ruler’. Furthermore, compared with ortholog LpxA structures, the purported oxyanion hole, formed by the backbone amide group of Gly139, displays a different conformation in P. aeruginosa LpxA, which suggests flexibility of this structural feature important for catalysis and the potential need for substrate-induced conformational change in catalysis.

Taken together, the three

structures provide valuable insights into P. aeruginosa LpxA catalysis and substrate specificity as well as templates for future inhibitor discovery.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 45

Pseudomonas aeruginosa is an opportunistic pathogen that can thrive in a variety of environments and infect many hosts including humans.1, 2 Serious infections can lead to death and are predominantly hospital-acquired.2 Furthermore, the pathogen has the ability to acquire resistance to multiple antibiotics through various mechanisms, making infections increasingly difficult to treat.3-6 Consequently, P. aeruginosa has entered the category of “superbugs” and is of major concern to clinicians, further highlighting the importance of identifying and studying new druggable targets.5-7 In P. aeruginosa, like most Gram-negative bacteria, lipopolysaccharide (LPS) is a major component of the outer membrane that protects the bacterium from its environment.8, 9 LPS is highly immunogenic and a major virulence factor, and as a result it is referred to as “endotoxin”.10-12 The LPS structure may differ slightly from one bacterium to another, but it is mainly comprised of three components; the O-antigen, the core oligosaccharide, and lipid A.9, 13, 14 Lipid A is a glucosamine-based phospholipid that anchors lipopolysaccharide to the outer monolayer of the outer membrane.8, 13, 15 In P. aeruginosa, lipid A is made out of a diglucosamine biphosphate backbone that is linked by O- and N- primary and secondary fatty acids.9, 13 This component is integral to the bacterium’s viability, and thus has become a desirable target for drug discovery.16 The pathway leading to the biosynthesis of lipid A has been well documented in E. coli,17, 18 and it is believed to be conserved across all Gram-negative species. The first step of

lipid

A

biosynthesis

involves

UDP-N-acetylglucosamine

(UDP-GlcNAc)

acyltransferase (LpxA), which catalyzes the transfer of a R-3-hydroxyacyl chain from the R-3-hydroxyacyl-acyl carrier protein (ACP) to the glucosamine 3-OH group of UDPGlcNAc (Fig. 1).16, 18 This first step is thermodynamically unfavorable (Keq of ~ 0.01 in E.


Page 5 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

coli)18, but LpxC, the second enzyme in the pathway, catalyzes the committed step through the thermodynamically favored and effectively irreversible deacetylation of UDP-3-O-(R-3-hydroxyacyl)-GlcNAc (Fig. 1).9, 19 The third step is catalyzed by LpxD through the addition of a second β-hydroxyacyl chain to make UDP-2,3-diacyl-GlcNAc (Fig. 1).9, 20 A series of additional steps result in the production of Lipid A that is attached to the core component of LPS (Fig. 1).8, 21 Previous studies have shown that LpxA and other enzymes involved in the LPS synthesis pathway provide validated targets for novel antibiotics.22-25 LpxA is a soluble cytoplasmic protein that forms a functional homotrimer.26, 27 It shares sequence homology with LpxD, which also forms a homotrimer, unlike LpxC which is a Zn2+-dependent monomeric enzyme that shares no sequence homology with other deacetylases.8 Previously solved crystal structures reveal that LpxA forms a left-handed parallel β-helix generated by ~30 hexapeptide repeats.28-34 Every turn in the β-helix is made up of a single hexapeptide.35 Conserved residues, identified previously through sequence alignments, assisted in the identification of the active site, which is located in a cleft formed by the interface of adjacent monomeric units of the trimer. The key catalytic residue, which is conserved across all species, is a histidine that has been proposed to deprotonate the 3-OH of the UDP-GlcNAc substrate, priming it for nucleophilic attack.36 For this to occur, the 3-OH of UDP-GlcNAc is positioned in the active site near His121 through the formation of multiple hydrogen bonds between the rest of UDP-GlcNAc and other residues. Finally, the acyl chain, carried by the ACP, docks into a hydrophobic cleft and is displaced from the ACP via nucleophilic attack by the 3-OH of UDP-GlcNAc and


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 45

subsequently attached to UDP-GlcNAc forming the product, UDP-3-O-(R-3hydroxyacyl)-GlcNAc.9 LpxA enzymes are selective of their substrate’s acyl chain length, but this selectivity varies across species, with some bacteria preferring longer chain hydrocarbons while other bacteria recognize only shorter chain hydrocarbons.37-39 In E. coli for example, LpxA catalyzes the transfer of R-3-hydroxymyristoyl, a 14C-long fatty acyl chain, from the R-3-hydroxymyristoyl-acyl carrier protein (ACP) to UDP-GlcNAc. In vitro, E. coli LpxA has a catalytic preference for this 14C-long chain over a 10C-long chain by a factor of ~1000.39 In L. interrogans, LpxA is selective for R-3-hydroxylauroyl, a 12C-long fatty acyl chain, which it attaches to the UDP-GlcNAc derivative, UDP-GlcNAc3N.28 In P. aeruginosa however, LpxA prefers R-3-hydroxydecanoyl, a 10C-long R-3-hydroacyl chain.39 Lipid A isolates from E. coli and P. aeruginosa have confirmed incorporation of their specific acyl chains.9, 12 The residues involved in restricting the length of the acyl chains in LpxA were first identified by Wyckoff et al in E. coli and P. aeruginosa, and named “hydrocarbon rulers”.39 These are the amino acids that act as precise measuring tools, allowing the incorporation of hydrocarbon chains of very specific lengths; longer chains are disallowed, while shorter chains do not provide sufficient hydrophobic interactions and therefore lose significant affinity. Structural studies have shown that in E. coli LpxA, His191 is implicated in restricting the chain length to be 14C,35 while in L. interrogans LpxA, structural studies have shown that Lys171 serves a similar purpose for the shorter 12C chain.28 In P. aeruginosa LpxA, Met169 has been functionally identified as the “hydrocarbon ruler”, denying longer chain hydrocarbons from binding.39 Furthermore, it has been shown that this selectivity can be switched in vitro and in vivo


Page 7 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

between E. coli LpxA and P. aeruginosa LpxA through reciprocal mutations of G173M and M169G respectively.39 Until now however, the structure of P. aeruginosa has been lacking, so a comprehensive understanding of how this selectivity is accomplished by Met169 was incomplete. Here we report the first structures of P. aeruginosa LpxA in three forms: apo protein at 1.81 Å, substrate complex with UDP-GlcNAc at 2.16 Å, and product complex with UDP3-0-(R-3-hydroxydecanoyl)-GlcNAc at 2.30 Å. These structures have unambiguously identified the active site and provide detailed information on the residues that orient UDP-GlcNAc for catalysis, including the role of His121 as a general base, and on how Met169 confers P. aeruginosa LpxA with its exceptional selectivity for R-3hydroxydecanoyl, the 10C-long hydrocarbon chain, thus serving as the “hydrocarbon ruler”. Such structural information is not only of significance for better understanding substrate recognition and catalysis in this class of enzymes, but also for aiding in the development of inhibitors using a structure-based approach.

EXPERIMENTAL PROCEDURES Materials All reagents and chromatography supplies were purchased from Fisher Scientific. Crystal screens were purchased from Qiagen. UDP-GlcNAc was purchased from SigmaAldrich. The LpxA product, UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc, was synthesized by the Alberta Research Council (Alberta, Canada) Purification of recombinant LpxA. The plasmid pET28b containing the N-terminal His-tagged P. aeruginosa LpxA sequence was transformed into Rosetta (DE3) pLysS


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 45

cells (Novagen). The cells were incubated in LB media supplemented with 35 µg/ml chloramphenicol and 50 µg/ml kanamycin at 37°C overnight. The overnight culture was then diluted into 1L LB media containing 35 µg/ml chloramphenicol and 50 µg/ml kanamycin at 1:100 and incubated at 37°C until the OD600 reached 0.6-0.8. The protein expression was initiated with 0.5 mM IPTG and incubation continued at 37°C for an additional 4 hours. The cells were harvested by centrifugation at 5,000g for 10 min. The pellet was resuspended in buffer A (20 mM Tris-HCl, pH 8.4, 250 mM NaCl, 20 mM imidazole, and 10 % glycerol). The cells were disrupted by sonication followed by centrifugation at 35,000g for 40 min to remove debris. The supernatant was then loaded to a HisTrap affinity column. The protein was eluted with a linear gradient of 500 mM imidazole. The fractions containing the tagged LpxA were pooled and concentrated. The sample was loaded to a HiLoad 16/60 Superdex 75 column for further purification in thrombin cleavage buffer (20 mM Tris pH 8.4, 150 mM NaCl and 10 % glycerol). The peak fractions containing the His-tagged LpxA were pooled, and the concentration of the protein was determined by OD280. The protease thrombin (Roche) was added at a ratio of one unit per mg of the protein. After overnight incubation at room temperature, the samples were then loaded onto a HisTrap column to remove any uncleaved protein. The flow-through was collected and concentrated, followed by gel filtration with the HiLoad 16/60 Superdex 75 column. The protein eluted at a peak consistent with the size of the trimeric form. The untagged LpxA was stored at -80°C at 14.3 mg/ml concentration in a buffer containing 20 mM potassium phosphate (pH 8.6) and 250 mM NaCl. The purity of the protein was determined by SDS-PAGE at >95%.


Page 9 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

LpxA Crystallization. Qiagen crystallization screens JCSG suites I-IV, AmSO4, MPD, Core I & II, were screened using the Phoenix nanodispenser, and 0.2 and 0.4 µl aliquots of protein solution (14.3 mg/ml), each with 0.2 µl well solution, were used to search for crystallization conditions. P. aeruginosa LpxA readily crystallized in many conditions, producing cuboidal crystals of poor X-ray diffraction quality. Crystals with an “almond”like morphology emerged in 0.1 M imidazole pH 8.0, 20% (w/v) PEG 1000, and 0.2 M calcium acetate, which diffracted to high resolution. The crystals appeared within 2 to 4 days and measured up to 0.1 mm in length. The apo crystal was soaked in crystallization buffer containing 25% glycerol and cryocooled in liquid nitrogen. The UDP-GlcNAc complex was obtained by transferring apo crystals into crystallization solution containing 50 mM UDP-GlcNAc and soaked overnight, while the UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc complex was obtained by transferring apo crystals into crystallization solution containing 10 mM UDP-3-O-(R-3hydroxydecanoyl)-GlcNAc and soaked over a period of three days. The crystals were then briefly soaked in crystallization buffer containing 25% glycerol and ligand, and then immediately cryocooled in liquid nitrogen. Data collection and structure determination. X-ray diffraction data for the apo and UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc complex LpxA crystals were collected at the SER-CAT ID beamline at the Advanced Photon Source (APS) at Argonne National Laboratory. X-ray diffraction data for the UDP-GlcNAc complex LpxA crystal was collected at the 8.3.1 beamline at the Advanced Light Source (ALS) at Lawrence Berkeley National Laboratory. For all three datasets, indexing, integration and scaling was done through HKL2000.40 The apo structure was solved via molecular replacement


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 45

in MolRep41 (CCP4 suite)42 using a homology model constructed through Modeller,43 which was based on the E. coli LpxA structure (PDB ID: 2AQ9).24 Model rebuilding was carried out using the program COOT,44 and refinement was performed using Phenix45 and Refmac546 (CCP4 suite).42 The complex structures were subsequently solved via molecular replacement using the apo structure as a template. Figures of protein structures were made using PyMOL (The PyMOL Molecular Graphics System, Version 1.3 Schrödinger, LLC). The omit map (Fig. 7C) was created in Phenix45 by omitting target regions and refining with simulated annealing to decrease model bias. Data collection and refinement statistics for all three structures are presented in Table 1. The protein coordinates have been deposited into the Protein Data Bank with accession code 5DEM for the apo protein, 5DEP for the substrate complex with UDP-GlcNAc, and 5DG3 for the product complex with UDP-3-0-(R-3-hydroxydecanoyl)-GlcNAc.

RESULTS AND DISCUSSION Apo structure. The P. aeruginosa LpxA apo structure was solved at 1.81 Å resolution and belongs to the P212121 space group (Table 1). Each asymmetric unit contains six monomers. Three of those monomers form a biologically relevant homotrimer through a non-crystallographic 3-fold symmetry (Fig. 2A). The remaining three monomers in the asymmetric unit also form functional homotrimers, but with monomers from adjacent asymmetric units. The conformation of each monomer is virtually identical and when superimposed to monomer A, they have an average RMSD value of 0.193 Å aligning an average of 1480 atoms. Superimposition of all the monomers in the biologically relevant trimer shows negligible structural variation, except at the more flexible loop region L1


Page 11 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(Glu66-Thr78) where differences are observed (Fig. 2B-C). P. aeruginosa LpxA contains high overall structural similarity to other previously solved LpxA structures, such as E. coli (54% sequence identity)30 and L. interrogans (41% sequence identity),28 yet it also displays differences in residue composition and side chain conformations, especially in flexible loop regions where backbone variations are observed (Fig. 2D). The amino acids in the apo structure were clearly resolved in the electron density map, with the exception of Met1 (chain A-F), and Ser2 (chain C). The structure of P. aeruginosa LpxA, similar to orthologs in other bacteria, is made up of two distinct domains; an N-terminal β-strand domain and a C-terminal α-helical domain (Fig. 2B-C). Ten β-helical coils form the typical LβH (Left-handed parallel beta-Helix) domain of the N-terminus, a motif common in enzymes with acyltransferase activity.20 Starting at Met1 and ending at Met193, this coiled motif consists mostly of hexapeptide repeats, giving rise to 29 distinct β-strands (Fig. 2B). Every three β-strands form one complete β-helical coil, creating the appearance of an equilateral triangle. Side chains in the coil hexapeptides alternate between those facing internally, and those facing externally. Those facing internally establish polar and non-polar interactions further rigidifying the structure. Two loops, L1 (Glu66-Thr78) and L2 (Arg96-Ala103), extend out of the LβH domain at coils C4-C5 and C5-C6, interrupting the β-helical motif (Fig. 2B-C). These loops show slight conformational variation among different monomers in the asymmetric unit (Fig. 2C), and display structural variations between orthologs in other bacteria (Fig. 2D). Immediately following the LβH domain, a series of four consecutive α-helices of different lengths and tilts are formed, starting at Asn194 and extending all the way to the


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 45

C-terminus (Fig. 2B-C). This α-helical domain shows some conformational variation among ortholog proteins from other bacterial and plant species (Fig. 2D).28-33 While P. aeruginosa LpxA is most similar to the E. coli LpxA (PDB ID: 1LXA),30 there seems to be some variation at the C-terminal end of the third α-helix, involving different tilts compared to bacterial orthologs A. baumannii (PDB ID: 4E6U),29 L. interrogans (PDB ID: 3HSQ),28 C. jejuni (PDB ID: 3R0S), B. thailandensis (PDB ID: 4EQY),32 H. pylori (PDB ID: 1J2Z),31 B. fragilis (PDB ID: 4R36),34 and plant ortholog A. thaliana (plant) (PDB ID: 3T57).33 The catalytic site in the biologically relevant trimer lies in the protein-protein interface where coils C6-C9 from adjacent monomers meet. Therefore, each biologically relevant trimer contains three catalytic sites. Even though there are some differences in active site residues among different orthologs, key catalytic residues, such as His121 in P. aeruginosa LpxA, are conserved across all species (Fig. 2D). This suggests the same mode of catalysis, but perhaps variation in ligand positioning and substrate specificity between orthologs.

LpxA/UDP-GlcNAc substrate complex. The P. aeruginosa LpxA/UDP-GlcNAc complex structure, obtained by soaking the substrate compound into the LpxA apo crystal, was solved at 2.16 Å resolution (Table 1). The resulting electron density unambiguously identified the presence and localization of six UDP-GlcNAc molecules within the asymmetric unit, three of which appear in the active site at the protein-protein interface. Only one of those three is at an active site contained within the asymmetric unit (Fig. 3), while the other two are at active sites formed at the crystal-packing interface


Page 13 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

together with monomers from adjacent asymmetric units. For these two ligands, applying crystallographic symmetry yields the complete active site and identifies all protein-ligand interactions, confirming that all three molecules bind in an identical manner to the active site. Electron density for the remaining UDP-GlcNAc molecules was observed outside the active site at three distinct sites. One of them is positioned near monomer C right at the cleft where α-helix 1 and the β-strand domain meet. Only two potential hydrogen bond interactions are possible for this binding pose, and the electron density for the GlcNAc moiety of the substrate is not well defined, indicating conformational disorder. This suggests that this binding pose is a crystal-packing artifact.

The other two

molecules however are positioned at the protein-protein interface, but at a site situated below the active site, on the N-terminal side of loop L1 (Fig. 4). One of these two molecules binds at the protein-protein interface within the asymmetric unit (below the active site) (Fig. 4), while the other is positioned at a protein-protein interface composed by monomers from adjacent asymmetric units, yet the binding poses of both molecules are nearly identical when crystallographic symmetry is applied. At this secondary site, the ligand establishes many favorable polar and non-polar interactions with residues from both monomers at the protein-protein interface within the biological homotrimer (Fig. 4). Cumulatively, these interactions are fewer than the interactions observed at the biologically relevant active site (Fig. 3A-B), but are still plentiful enough to possibly contribute to binding in vivo, increasing the local concentrations of the substrate near the protein active site. The residues interacting with the ligand are not conserved but it would be of interest to test whether this could be a biologically relevant secondary site unique to P. aeruginosa LpxA, or is simply a crystal-packing artifact. However, the focus


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 45

hereinafter will be on the active site formed within the asymmetric unit at the proteinprotein interface between monomers A and B (Fig. 3). UDP-GlcNAc interacts with the active site mostly through polar interactions with residues from both monomers A and B (Fig. 3A). The uracil moiety and certain atoms of the diphosphate moiety hydrogen bond to monomer A (Fig. 3A), while the ribose, GlcNAc, and other atoms of the diphosphate moieties interact with monomer B (Fig. 3AB). On the one end of the UDP-GlcNAc molecule, the 3-N and 4-O atoms of uracil are within hydrogen bond distance to Asn194/Oδ1 and Asn194/Nδ2 (chain A) respectively. The 4-O atom of uracil is also within hydrogen bond distance to Arg201/Nη2 (Fig. 3A). The remaining non-polar atoms of the uracil ring appear to lie against the aromatic side chain of Phe166 (chain A), which also forms van der Waals contacts with two carbon atoms on the ribose ring. Van der Waals contacts also form between the uracil ring and Ile148 on one side, as well as with the apolar side chain atoms of Glu196 on the other side. The 3’-OH of the ribose moiety hydrogen bonds with His156/Nε2 (chain B), while the remaining polar atoms of the ribose appear to interact only with the solvent. Arg200/Nη2 (chain A) interacts with the linker-O atom of α-phosphate and Arg200/Nη1 with an O atom of the β-phosphate (Fig. 3A), while an α-phosphate O atom is within hydrogen bond distance to Gln157/Nε2 (chain B) (Fig. 3A). Meanwhile, the GlcNAc moiety appears to implicate an abundance of hydrogen bonds with residues of chain B (Fig. 3A-B). The GlcNAc 6-OH hydrogen bonds with a series of polar atoms, such as Tyr158/Oη, His140/Nε2, and Lys72/Nζ (Fig. 3B). The 3-OH group of GlcNAc is within hydrogen bond distance to the proposed catalytic base His121/Nε2 (chain B), which is believed to function by deprotonating 3-OH and priming it for nucleophilic attack (Fig.


Page 15 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

3B). The 4-OH group of GlcNAc is also within hydrogen bond distance to His121/Nε2, but at a less favorable angle than the 3-OH group. Lastly, the acetyl O atom hydrogen bonds with the backbone Leu71/N atom (Fig. 3B). The apolar atoms of the Asp70 side chain form van der Waals contacts with the apolar atoms of the GlcNAc moiety, while the Leu71 side chain appears within van der Waals distance to the apolar side of the acetyl group (Fig. 3A). Comparison of the LpxA apo structure to the LpxA/UDP-GlcNAc complex shows the conformational changes that are induced upon binding the substrate (Fig. 3C). Gln157 (chain B) has flipped away from the pocket to make room for the ligand and to hydrogen bond with the α-phosphate, while Lys72 (chain B) moves in to interact with the GlcNAc moiety (Fig. 3C). Asp70 (chain B) also moves away from the pocket to accommodate the ligand, and allow the non-polar atoms of its side chain to provide a hydrophobic surface for the apolar atoms of the GlcNAc moiety (Fig. 3C). On chain A, both Arg200 and Arg201 adopt alternative conformations in the complex structure, by moving closer to interact with UDP-GlcNAc (Fig. 3C). Finally, Glu196 moves away from the pocket to accommodate the ligand and to also provide additional apolar surface for the uridine moiety (Fig. 3C). Due to these interactions between UDP-GlcNAc and chain A, the entire Asn194-Ser205 backbone segment of chain A is pulled closer towards the ligand in the complex structure compared to the apo form (Fig. 3C). Currently in the PDB, there is an E. coli ortholog LpxA structure complexed with UDP-GlcNAc (PDB ID: 2JF3).47 The E. coli/UDP-GlcNAc complex structure contains one monomer per asymmetric unit. By applying crystallographic symmetry operations and superimposing the E. coli complex to the P. aeruginosa LpxA complex structure,


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 45

direct comparison of the binding pose of UDP-GlcNAc between these two orthologs demonstrates both significant differences and similarities in UDP-GlcNAc binding (Fig. 3D). The biggest difference lies in the positioning of the uridine moiety, which is flipped around the phosphate-O linker bond, when comparing the two complexes (Fig. 3D). This conformational difference may have resulted from variations in the interactions between the protein and the substrate.

Some potential hydrophobic interactions, like the ones

observed between UDP-GlcNAc and Phe166 (chain A), are not present in the E. coli complex, since the equivalent Met170 would be unable to establish the same interactions with both the uracil and ribose groups. Also, even though His156 (chain B), which interacts with the ribose moiety in P. aeruginosa LpxA, is conserved between these two orthologs, in the E. coli complex the equivalent His160 does not interact with the ribose but instead it is flipped away from the ligand and occupies a space normally held by Met169 in the P. aeruginosa LpxA (Fig. 3D). Another reason why the alternative uridine moiety conformation is observed may be due to the slightly wider binding site that accommodates the uridine moiety in the E. coli LpxA. This difference is especially noticeable when the E. coli and P. aeruginosa orthologs are superimposed, where chain A in P. aeruginosa is shifted closer towards the ligand making the pocket narrower (Fig. 3D), and as a consequence the alternative uridine moiety conformation would have been obstructed by residue Ile148 (Ile152 in E. coli). On the other hand, while the position of the α-phosphate is different in these two binding poses, the position of the β-phosphate is quite similar, and even more so is the positioning of the GlcNAc moiety. Therefore, the hydrogen bonding network between the GlcNAc moieties in these two complexes is very


Page 17 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

similar. An exception is that Tyr158 in P. aeruginosa LpxA is Phe162 in E. coli LpxA (Fig. 3D), and is thus incapable of hydrogen bonding to 6-OH of GlcNAc. In addition to the E. coli LpxA-UDP-GlcNAc complex, there is another most recent UDP-GlcNAc complex crystal structure with the B. fragilis LpxA ortholog (PDB ID: 4R37).34 In the B. fragilis LpxA complex, the substrate shows a binding pose that is nearly identical to the P. aeruginosa complex, while the binding site, like in the P. aeruginosa LpxA complex, is also narrower than the E. coli structure. Cumulatively, the structural information obtained from all three orthologs complex structures further supports that the mechanism of catalysis is conserved.

LpxA/UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc

product

complex.

The

P.

aeruginosa LpxA/UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc complex structure was solved at 2.30 Å resolution (Table 1). Electron density for the product was observed in four of the six active sites associated with the asymmetric unit, compared to only three active sites occupied in the LpxA/UDP-GlcNAc complex structure. Extensive hydrophobic contacts of the product within the acyl chain binding site may help explain why the substrate UDP-GlcNAc was found in only three active sites compared to the four products found in the product complex. Addition of the acyl chain also appears to prevent the product from being accommodated at the non-catalytic sites where the substrate was found. Superimposition of the four molecules shows negligible conformational variation, and thus the focus hereafter will be at the active site formed between chain A and B (Fig. 5).


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 45

The unbiased Fo-Fc map unambiguously identifies the binding pose of UDP-3-O-(R-3hydroxydecanoyl)-GlcNAc at the A-B chain interface, which shares striking similarities with the substrate complex in this region (Fig. 5A). The UDP-GlcNAc moiety is engaged in the same hydrogen bonding network as seen in the substrate complex (Fig. 3A, 5A). The acylated 3-O of the GlcNAc moiety is positioned similarly to the GlcNAc 3-OH of the substrate, and the acyl chain extends from there into the acyl chain binding cleft (Fig. 5A-B). His118/Nε2 appears within hydrogen bond distance to the acyl chain hydroxyl, which might help anchor the chain into the hydrophobic cavity (Fig. 5B). The acyl chain establishes multiple hydrophobic contacts with the protein. Apolar atoms of the acyl chain complement the hydrophobic tunnel formed by chain A residues Val132, Tyr152, Met169, the non-polar side chain atoms of Asn133, and the backbone of Gly151, as well as chain B residues Ala136, Ala138, and Leu154 (Fig. 5B). Structures of E. coli and L. interrogans LpxA complexed to their acylated UDPGlcNAc products have been previously solved; the E. coli LpxA is in complex with UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc, a 14C-long chain product (PDB ID: 2QIA),35 and the L. interrogans LpxA is in complex with its 12C-long acyl chain product (PDB ID: 3I3X).28 Superimposition of our P. aeruginosa LpxA complex structure to the E. coli LpxA complex shows that UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc binds to E. coli LpxA in an orientation similar to how UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc binds to P. aeruginosa LpxA (Fig. 5C), with only minor differences.

The most

noticeable difference is observed in the slight shift of the uridine moiety towards chain A in E. coli, probably as a result of the slightly wider binding site in E. coli LpxA. Even though His160 in E. coli LpxA adopts the same conformation as His156 in P. aeruginosa


Page 19 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

LpxA, it is not within hydrogen bond distance to the ribose 3’-OH of the product (Fig. 5C). Superimposition to the previously solved L. interrogans LpxA/product complex structure also suggests that acylated products bind similarly in both L. interrogans LpxA and P. aeruginosa LpxA (Fig. 5D). Yet, different orientations for the 2-N acetyl group are observed when comparing the GlcNAc moiety in P. aeruginosa and the GlcNAc3N moiety in L. interrogans (Fig. 5D). This difference may be due to the presence of Phe147, in place of Ile148 in P. aeruginosa, with the former capable of reaching and establishing hydrophobic interactions with the apolar side of the 2-N acetyl group (Fig. 5D). Additionally, Leu70 in L. interrogans has adopted an alternative conformation from the equivalent Leu71 in P. aeruginosa, and would sterically clash with the 2-N acetyl conformation observed in the P. aeruginosa complex structure (Fig. 5D). Another small difference is also observed in the orientation of the GlcNAc 6-OH, probably due to the presence of Gly71 instead of Lys in the L. interrogans ortholog, which is incapable of interacting with the GlcNAc 6-OH group (Fig. 5D). The major difference between the three orthologs however lies in the hydrophobic tunnel that accommodates the acyl chains of the acylated products and restricts their lengths (Fig. 6). In lipid A biosynthesis, precise incorporation of specific hydrocarbon length acyl chains is highly conserved within a species, but divergent between different species. The term “hydrocarbon ruler” was first coined by Wyckoff et al, in describing residues that can act as precise measuring tools in the E. coli and P. aeruginosa LpxA.39 Such residues would disallow longer chains from binding due to steric clashes, and would not provide sufficient hydrophobic interactions to shorter chains and therefore lose significant affinity. In E. coli LpxA, both structural and functional work support His191


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 45

acting as the “hydrocarbon ruler”, while in L. interrogans LpxA, it is Lys171. Mutational studies had suggested Met169 to serve the same role in P. aeruginosa LpxA and also showed how reciprocal mutations can completely reverse this selectivity between E. coli and P. aeruginosa.39 The E. coli LpxA G173M mutant lost its preference for the 14Clong substrate while gaining activity for the 10C-long substrate – this resulted in a 106fold selectivity switch from the C14 to the C10 substrate. Conversely, it was also shown that the M169G mutation in P. aeruginosa LpxA switched its specificity to the 14C-long product by 106-fold.39 In E. coli LpxA, Gly173 lies in the same position as Met169 in P. aeruginosa LpxA. The function of Met169 is now evident in our complex structure and our results provide detailed structural information and support all previous functional findings. The LpxA/product complex reveals that product molecules with hydroxyacyl chains longer than 10 hydrocarbons would sterically clash with residue Met169, while shorter hydrocarbon chains would have reduced hydrophobic interactions with the pocket and therefore result in decreased affinity (Fig. 5A-B, 6A). Superimposition of the P. aeruginosa LpxA with E. coli and L. interrogans LpxA complexes confirms that 14 and 12 hydrocarbon-long chains would clash with Met169 in the P. aeruginosa LpxA. Similarly, in the L. interrogans LpxA, a hydroxyacyl chain longer than a 12 hydrocarbonlong acyl chain would clash with Lys171 (Fig. 6B), and in the E. coli LpxA, anything longer than a 14 hydrocarbon-long acyl chain would clash with His191 (Fig. 6C). Interestingly, in these two LpxA structures, the “hydrocarbon ruler” (Lys171 and His191 respectively) resides on the same monomer as the catalytic histidine, whereas in P. aeruginosa, Met169 is provided in trans from a different monomer (Fig. 5A-B, 6A) in


Page 21 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

comparison to His121. Also, since P. aeruginosa LpxA has a Pro187 in the place of His191 from E. coli LpxA, no steric hindrance with the acyl chain would be expected at that site. It would therefore be interesting to test what is the longest hydrocarbon chain that could fit in the P. aeruginosa LpxA M169G mutant. It has been previously proposed that the backbone amide group of a conserved Gly residue (Gly139 in P. aeruginosa LpxA) acts as the oxyanion hole by stabilizing the transition state during acylation of UDP-GlcNAc.28,

35

For that to occur in the P.

aeruginosa LpxA, the corresponding amide linker should be positioned in a way such that the carbonyl from the preceding Ala138 residue is orientated away from the active site and the backbone N of Gly139 is directed towards the active site, as observed in the E. coli and L. interrogans complex structures (backbone NH of residues Gly143 and Gly138, respectively) (Fig. 7A-B), as well as all the other ortholog structures.28, 29, 31-35 However, this conformation is not observed in the P. aeruginosa LpxA structure (Fig. 7). Instead, the backbone N of Gly139 flips away from the active site, and the backbone carbonyl of Ala138 is oriented towards the ligand. Simulated-annealing omit maps of this region in the apo (1.81 Å) (Fig. 7C), substrate-complex (2.16 Å), and productcomplex (2.30 Å) all support the same conformational arrangement, which is most evident in the apo structure due to the higher resolution (Fig. 7C). It also appears that the acyl chain carbonyl of UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc, which would normally interact with the oxyanion hole during the reaction, is directed away from the protein and towards the solvent (Fig. 7A-B). In the E. coli (Fig. 7A) and L. interrogans (Fig. 7B) complex structures, the acyl chain carbonyl is positioned approximately within


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 45

hydrogen bond distance (3.3 Å) to the respective Gly residue, consistent with the role of the conserved Gly as the oxyanion hole (Fig. 7A-B). The favorable Gly backbone conformation, opposite to the one in our structures, has been observed not only in every other LpxA ortholog structure, but also in every LpxD structure solved to date.48-51 The alternative conformation in our structures may have been due to increased flexibility of the Gly139 residue in P. aeruginosa LpxA. In E. coli LpxA, the peptide bond between Ala142 and Gly143, equivalent to Ala138 and Gly139 in P. aeruginosa, is stacked on top of the amide linkage between Ala124 and the catalytic histidine, His125. If the amide group adopts the same conformation as observed in P. aeruginosa LpxA, the backbone carbonyl group of Ala142 would clash with the Cβ atom of Ala124 in E. coli LpxA. In addition, the area that accommodates the acyl linkage appears to be wider in P. aeruginosa LpxA, due to a substitution of Pro69 for the equivalent Gln73 in E. coli (Fig. 7A). The distance between Gly139/Cα and Pro69/Cα is 10.8 Å in P. aeruginosa LpxA, in comparison to 8.3 Å between the corresponding atoms in E. coli LpxA. As a result, in the P. aeruginosa LpxA product complex, the acyl linkage is positioned further away from Gly139 in order to form favorable interactions with Pro69. It is possible that the favorable Gly139 conformation as observed in other LpxA structures may be adopted by P. aeruginosa LpxA during catalysis, especially taking into consideration that our purified P. aeruginosa LpxA is biologically active (as tested in activity assays with the UDP-GlcNAc and acylated-ACP substrates - data not shown). In a recent E. coli LpxD-ACP complex (PDB ID: 4IHF),52 the carbonyl group of the acyl chain, still attached to the 4’-phosphopantheteine group (4’-PPT) on the carrier protein, is pointed at the N group of the equivalent glycine residue and forms a hydrogen


Page 23 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

bond with a length of 3.1 Å, in comparison to 3.3 Å for the equivalent hydrogen bond in the E. coli LpxA product complex. This suggests that functional groups from ACP, such as 4’-PPT, can help position the acyl-chain carbonyl group for catalysis by interacting with the LpxA/LpxD active site. By placing the acyl chain carbonyl group close to Gly139, the binding of ACP may also induce the necessary conformational change to create the oxyanion hole.

CONCLUSION P. aeruginosa, a pathogenic Gram-negative bacteria with a wide range of antibiotic resistance mechanisms, has been causing increasing public health concerns.4, 5 Due to its importance in LPS synthesis, LpxA is a valuable target in drug discovery against P. aeruginosa.7,

16

Even though eight ortholog LpxA structures had previously been

solved,28-34 only three were in complex with glucosamine ligands and the P. aeruginosa LpxA structure was still lacking. Our X-ray structures reveal for the first time the architecture of P. aeruginosa LpxA and its detailed conformations in three different states. They provide valuable information on the catalytic mechanism and substrate specificity of this protein, highlighting both the similarities to orthologs from other bacteria as well as the unique features of a “hydrocarbon ruler” and a flexible oxyanion hole. This information is also invaluable in employing a structure based drug discovery approach that can lead to the discovery of new inhibitors against P. aeruginosa LpxA. Both the highly positively-charged electrostatic features of the active site and the hydrophobic nature of the hydrocarbon chain cavity can be exploited simultaneously in designing molecules that will compete with the acylation of UDP-GlcNAc.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 45

Furthermore, considering the restricted depth of the hydrocarbon cavity governed by Met169, and the unique composition of certain active site residues, the design of potent molecules specific to the P. aeruginosa strain is indeed promising.

AKNOWLEDGEMENTS We would like to thank Dr. Ruslan Sanishvili from GM-CA CAT at the Advanced Photon Source (APS) for help with crystal analysis, and the beamline scientists and staff at SER-CAT at APS for support with data collection. We thank Derek Nichols and Kyle Kroeck for reading the manuscript.


Page 25 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

REFERENCES 1. Tummler, B., Wiehlmann, L., Klockgether, J., and Cramer, N. (2014) Advances in understanding Pseudomonas, F1000Prime Rep 6, 9. 2. Morita, Y., Tomida, J., and Kawamura, Y. (2014) Responses of Pseudomonas aeruginosa to antimicrobials, Front Microbiol 4, 422. 3. Bonomo, R. A., and Szabo, D. (2006) Mechanisms of multidrug resistance in Acinetobacter species and Pseudomonas aeruginosa, Clin Infect Dis 43 Suppl 2, S4956. 4. Lister, P. D., Wolter, D. J., and Hanson, N. D. (2009) Antibacterial-resistant Pseudomonas aeruginosa: clinical impact and complex regulation of chromosomally encoded resistance mechanisms, Clin Microbiol Rev 22, 582-610. 5. Schurek K. N., Breidenstein E. B. M., Hancock R. E. W. (2012) Pseudomonas aeruginosa: a persistent pathogen in cystic fibrosis and hospital-associated infections, in Antibiotic Discovery and Development 1st ed., pp 679–715, Springer, US. 6. Wolska, K., Kot, B., Piechota, M., and Frankowska, A. (2013) [Resistance of Pseudomonas aeruginosa to antibiotics], Postepy Hig Med Dosw (Online) 67, 13001311. 7. Rusmini, R., Vecchietti, D., Macchi, R., Vidal-Aroca, F., and Bertoni, G. (2014) A shotgun antisense approach to the identification of novel essential genes in Pseudomonas aeruginosa, BMC Microbiol 14, 24. 8. Wang, X., and Quinn, P. J. (2010) Lipopolysaccharide: Biosynthetic pathway and structure modification, Prog Lipid Res 49, 97-107.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 45

9. King, J. D., Kocincova, D., Westman, E. L., and Lam, J. S. (2009) Review: Lipopolysaccharide biosynthesis in Pseudomonas aeruginosa, Innate Immun 15, 261312. 10. Heine, H., Rietschel, E. T., and Ulmer, A. J. (2001) The biology of endotoxin, Mol Biotechnol 19, 279-296. 11. Raetz, C. R. (1993) Bacterial endotoxins: extraordinary lipids that activate eucaryotic signal transduction, J Bacteriol 175, 5745-5753. 12. Raetz, C. R., and Whitfield, C. (2002) Lipopolysaccharide endotoxins, Annu Rev Biochem 71, 635-700. 13. Lam, J. S., Taylor, V. L., Islam, S. T., Hao, Y., and Kocincova, D. (2011) Genetic and Functional Diversity of Pseudomonas aeruginosa Lipopolysaccharide, Front Microbiol 2, 118. 14. Kocincova, D., and Lam, J. S. (2011) Structural diversity of the core oligosaccharide domain of Pseudomonas aeruginosa lipopolysaccharide, Biochemistry (Mosc) 76, 755760. 15. Raetz, C. R., Reynolds, C. M., Trent, M. S., and Bishop, R. E. (2007) Lipid A modification systems in gram-negative bacteria, Annu Rev Biochem 76, 295-329. 16. Galloway, S. M., and Raetz, C. R. (1990) A mutant of Escherichia coli defective in the first step of endotoxin biosynthesis, J Biol Chem 265, 6394-6402. 17. Anderson, M. S., and Raetz, C. R. (1987) Biosynthesis of lipid A precursors in Escherichia

coli.

A

cytoplasmic

acyltransferase

that

converts

UDP-N-

acetylglucosamine to UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine, J Biol Chem 262, 5159-5169.


Page 27 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

18. Anderson, M. S., Bull, H. G., Galloway, S. M., Kelly, T. M., Mohan, S., Radika, K., and Raetz, C. R. (1993) UDP-N-acetylglucosamine acyltransferase of Escherichia coli. The first step of endotoxin biosynthesis is thermodynamically unfavorable, J Biol Chem 268, 19858-19865. 19. Young, K., Silver, L. L., Bramhill, D., Cameron, P., Eveland, S. S., Raetz, C. R., Hyland, S. A., and Anderson, M. S. (1995) The envA permeability/cell division gene of Escherichia coli encodes the second enzyme of lipid A biosynthesis. UDP-3-O-(R3-hydroxymyristoyl)-N-acetylglucosamine deacetylase, J Biol Chem 270, 3038430391. 20. Bartling, C. M., and Raetz, C. R. (2008) Steady-state kinetics and mechanism of LpxD, the N-acyltransferase of lipid A biosynthesis, Biochemistry 47, 5290-5302. 21. Raetz, C. R., Guan, Z., Ingram, B. O., Six, D. A., Song, F., Wang, X., and Zhao, J. (2009) Discovery of new biosynthetic pathways: the lipid A story, J Lipid Res 50 Suppl, S103-108. 22. Vaara, M. (1996) Lipid A: target for antibacterial drugs, Science 274, 939-940. 23. Onishi, H. R., Pelak, B. A., Gerckens, L. S., Silver, L. L., Kahan, F. M., Chen, M. H., Patchett, A. A., Galloway, S. M., Hyland, S. A., Anderson, M. S., and Raetz, C. R. (1996) Antibacterial agents that inhibit lipid A biosynthesis, Science 274, 980-982. 24. Williams, A. H., Immormino, R. M., Gewirth, D. T., and Raetz, C. R. (2006) Structure of UDP-N-acetylglucosamine acyltransferase with a bound antibacterial pentadecapeptide, Proc Natl Acad Sci U S A 103, 10877-10882. 25. Jenkins, R. J., and Dotson, G. D. (2012) Dual targeting antibacterial peptide inhibitor of early lipid A biosynthesis, ACS Chem Biol 7, 1170-1177.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 45

26. Dotson, G. D., Kaltashov, I. A., Cotter, R. J., and Raetz, C. R. (1998) Expression cloning of a Pseudomonas gene encoding a hydroxydecanoyl-acyl carrier proteindependent UDP-GlcNAc acyltransferase, J Bacteriol 180, 330-337. 27. Pfitzner, U., Raetz, C. R., and Roderick, S. L. (1995) Crystallization of UDP-Nacetylglucosamine O-acyltransferase from Escherichia coli, Proteins 22, 191-192. 28. Robins, L. I., Williams, A. H., and Raetz, C. R. (2009) Structural basis for the sugar nucleotide and acyl-chain selectivity of Leptospira interrogans LpxA, Biochemistry 48, 6191-6201. 29. Badger, J., Chie-Leon, B., Logan, C., Sridhar, V., Sankaran, B., Zwart, P. H., and Nienaber, V. (2012) Structure determination of LpxA from the lipopolysaccharidesynthesis pathway of Acinetobacter baumannii, Acta Crystallogr Sect F Struct Biol Cryst Commun 68, 1477-1481. 30. Raetz, C. R., and Roderick, S. L. (1995) A left-handed parallel beta helix in the structure of UDP-N-acetylglucosamine acyltransferase, Science 270, 997-1000. 31. Lee, B. I., and Suh, S. W. (2003) Crystal structure of UDP-N-acetylglucosamine acyltransferase from Helicobacter pylori, Proteins 53, 772-774. 32. Baugh, L., Gallagher, L. A., Patrapuvich, R., Clifton, M. C., Gardberg, A. S., Edwards, T. E., Armour, B., Begley, D. W., Dieterich, S. H., Dranow, D. M., Abendroth, J., Fairman, J. W., Fox, D., 3rd, Staker, B. L., Phan, I., Gillespie, A., Choi, R., Nakazawa-Hewitt, S., Nguyen, M. T., Napuli, A., Barrett, L., Buchko, G. W., Stacy, R., Myler, P. J., Stewart, L. J., Manoil, C., and Van Voorhis, W. C. (2013) Combining functional and structural genomics to sample the essential Burkholderia structome, PLoS One 8, e53851.


Page 29 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

33. Joo, S. H., Chung, H. S., Raetz, C. R., and Garrett, T. A. (2012) Activity and crystal structure

of

Arabidopsis

thaliana


acyltransferase,

Biochemistry 51, 4322-4330. 34. Ngo, A., Fong, K. T., Cox, D. L., Chen, X., and Fisher, A. J. (2015) Structures of Bacteroides fragilis uridine 5'-diphosphate-N-acetylglucosamine (UDP-GlcNAc) acyltransferase (BfLpxA), Acta Crystallogr D Biol Crystallogr 71, 1068-1076. 35. Williams, A. H., and Raetz, C. R. (2007) Structural basis for the acyl chain selectivity and mechanism of UDP-N-acetylglucosamine acyltransferase, Proc Natl Acad Sci U S A 104, 13543-13550. 36. Wyckoff, T. J., and Raetz, C. R. (1999) The active site of Escherichia coli UDP-Nacetylglucosamine

acyltransferase.

Chemical

modification

and

site-directed

mutagenesis, J Biol Chem 274, 27047-27055. 37. Williamson, J. M., Anderson, M. S., and Raetz, C. R. (1991) Acyl-acyl carrier protein specificity of UDP-GlcNAc acyltransferases from gram-negative bacteria: relationship to lipid A structure, J Bacteriol 173, 3591-3596. 38. Odegaard, T. J., Kaltashov, I. A., Cotter, R. J., Steeghs, L., van der Ley, P., Khan, S., Maskell, D. J., and Raetz, C. R. (1997) Shortened hydroxyacyl chains on lipid A of Escherichia

coli

cells

expressing

a

foreign


O-

acyltransferase, J Biol Chem 272, 19688-19696. 39. Wyckoff, T. J., Lin, S., Cotter, R. J., Dotson, G. D., and Raetz, C. R. (1998) Hydrocarbon rulers in UDP-N-acetylglucosamine acyltransferases, J Biol Chem 273, 32369-32372.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 45

40. Otwinowski, Z. M., W. (1997) Processing of X-ray Diffraction Data Collected in Oscillation Mode, in Methods in Enzymology, pp 307-326, C.W. Carter, Jr. & R. M. Sweet, Eds., New York. 41. Vagin, A., and Teplyakov, A. (2010) Molecular replacement with MOLREP, Acta Crystallogr D Biol Crystallogr 66, 22-25. 42. Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S. (2011) Overview of the CCP4 suite and current developments, Acta Crystallogr D Biol Crystallogr 67, 235-242. 43. Eswar, N., Webb, B., Marti-Renom, M. A., Madhusudhan, M. S., Eramian, D., Shen, M. Y., Pieper, U., and Sali, A. (2007) Comparative protein structure modeling using MODELLER, Curr Protoc Protein Sci Chapter 2, Unit 2 9. 44. Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics, Acta Crystallogr D Biol Crystallogr 60, 2126-2132. 45. Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C., and Zwart, P. H. (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution, Acta Crystallogr D Biol Crystallogr 66, 213-221. 46. Vagin, A. A., Steiner, R. A., Lebedev, A. A., Potterton, L., McNicholas, S., Long, F., and Murshudov, G. N. (2004) REFMAC5 dictionary: organization of prior chemical


Page 31 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

knowledge and guidelines for its use, Acta Crystallogr D Biol Crystallogr 60, 21842195. 47. Ulaganathan, V., Buetow, L., and Hunter, W. N. (2007) Nucleotide substrate recognition by UDP-N-acetylglucosamine acyltransferase (LpxA) in the first step of lipid A biosynthesis, J Mol Biol 369, 305-312. 48. Bartling, C. M., and Raetz, C. R. (2009) Crystal structure and acyl chain selectivity of Escherichia coli LpxD, the N-acyltransferase of lipid A biosynthesis, Biochemistry 48, 8672-8683. 49. Buetow, L., Smith, T. K., Dawson, A., Fyffe, S., and Hunter, W. N. (2007) Structure and reactivity of LpxD, the N-acyltransferase of lipid A biosynthesis, Proc Natl Acad Sci U S A 104, 4321-4326. 50. Badger, J., Chie-Leon, B., Logan, C., Sridhar, V., Sankaran, B., Zwart, P. H., and Nienaber, V. (2011) The structure of LpxD from Pseudomonas aeruginosa at 1.3 A resolution, Acta Crystallogr Sect F Struct Biol Cryst Commun 67, 749-752. 51. Badger, J., Chie-Leon, B., Logan, C., Sridhar, V., Sankaran, B., Zwart, P. H., and Nienaber, V. (2013) Structure determination of LpxD from the lipopolysaccharidesynthesis pathway of Acinetobacter baumannii, Acta Crystallogr Sect F Struct Biol Cryst Commun 69, 6-9. 52. Masoudi, A., Raetz, C. R., Zhou, P., and Pemble, C. W. IV. (2014) Chasing acyl carrier protein through a catalytic cycle of lipid A production, Nature 505, 422-426.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 45

Table 1. Data collection and refinement statistics LpxA apo

LpxA/UDP-GlcNAc

LpxA/UDP-3-O-(R-3hydroxydecanoyl)GlcNAc

Space group

P212121

P212121

P212121

Cell dimensions: a, b, c (Å)

79.03, 82.18, 220.51

80.04, 83.56, 222.43

79.96, 83.33, 220.37

Resolution range (Å)

110.26-1.81 (1.84-1.81)*

111.21-2.16 (2.20-2.16)*

110.18-2.30 (2.34-2.30)*

/

38.5 (2.3)*

13.1 (2.1)*

26.3 (2.2)*

Completeness (%)

98.8 (97.8)*

98.4 (99.5)*

98.5 (98.1)*

Rmerge

0.036 (0.770) *

0.066 (0.755) *

0.051 (0.580) *

Redundancy

7.5 (7.6) *

6.2 (6.3) *

6.4 (6.3) *

Number of reflections used

122365

75873

57656

Rwork

0.184

0.177

0.193

Rfree

0.211

0.226

0.245

11832/5/816

11800/254/411

11663/209/231

Bond lengths (Å)

0.0117

0.0115

0.0186

Bond angles (o)

1.3973

1.5581

1.8159

24.1/39.8/30.3

33.9/47.5/36.4

32.4/29.5/28.3

Most favored (%)

87.4

87.1

86.7

Additional allowed (%)

11.7

12.3

12.1

Generously allowed (%)

1.0

0.7

1.2

PDB code

5DEM

5DEP

5DG3

Data collection

Refinement

Number of atoms Protein/ ligands/ water Root mean square deviations

Average B-factor (Å2) Protein/ ligands/ water Ramachandran plot statistics**

*Numbers in parentheses are for the highest resolution shell. **Ramachandran plot statistics were calculated with ProCheck V 3.4.4 from the PDB validation server


Page 33 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Figure Legends Figure 1. P. aeruginosa LpxA catalyzes the first and reversible step of lipid A biosynthesis. LpxA catalyzes the reversible transfer of a hydroxyacyl chain from the ACP protein to the 3-OH position of UDP-GlcNAc. LpxC deacetylates UDP-3-O-(R-3hydroxydecanoil)-GlcNAc and LpxD catalyzes the transfer of a second hydroxyacyl chain from ACP to the 2-NH2 position. A series of additional enzymes carry out the remaining steps that lead to the biosynthesis of Lipid A. Figure 2. P. aeruginosa LpxA apo crystal structure. A. The biologically relevant homotrimer contained within the assymetric unit seen from two side views and one top view. B. Diagram of the secondary structure. C. Superimposition of all three monomers from the biologically relevant homotrimer, showing loops L1 and L2. D. P. aeruginosa LpxA monomer superimposed onto eight ortholog LpxA monomeric structures (E. coli (PDB ID: 1LXA), B. thailandensis (PDB ID: 4EQY), A. baumannii (PDB ID: 4E6U), L. interrogans (PDB ID: 3HSQ), H. pylori (PDB ID: 1J2Z), A. thaliana (PDB ID: 3T57), C. jejuni (PDB ID: 3ROS), and B. fragilis (PDB ID: 4R36)). Superimposition shows high structural similarity, the highly conserved catalytic histidine, and some variation in flexible loop regions and α-helical domains. Figure 3. UDP-GlcNAc bound to P. aeruginosa LpxA active site. Monomers A and B of the P. aeruginosa complex are colored in cyan and green respectively, whereas the substrate is in yellow. Hydrogen bonding interactions in the active site at the dimer interface are represented by dotted lines. A. The unbiased Fo-Fc map (contoured at 3.0 σ) unambiguously identifies the binding pose of UDP-GlcNAc in the active site. Uracil and


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 45

phosphate moieties interact with monomer A (cyan), while ribose, phosphate and glucosamine moieties interact with monomer B (green). B. Details of the interactions between the glucosamine moiety and monomer B. The 3-OH group is within hydrogen bond distance to catalytic residue His121. C. Comparison of P. aeruginosa LpxA/UDPGlcNAc complex to apo P. aeruginosa Lpxa (light grey) showing only residue side chains that demonstrate conformational changes upon binding of UDP-GlcNAc. D. Comparison of P. aeruginosa LpxA/UDP-GlcNAc complex to E. coli LpxA/UDPGlcNAc complex (light grey) showing the residues that orchestrate UDP-GlcNAc binding on E. coli LpxA. His156 and Tyr158 of P. aeruginosa LpxA are also shown but not labelled (P. aeruginosa ligand (yellow)/E. coli ligand (light pink)). Residue labels are based on E. coli LpxA sequence. Figure 4. UDP-GlcNAc observed at a secondary binding site at the dimer interface. Monomers A and B of the P. aeruginosa complex are colored in cyan and green respectively, whereas the substrates are in yellow. A UDP-GlcNAc molecule is bound outside the active site on the N-terminal side of loop L1 at the same dimer interface. Potential hydrogen bonds between UDP-GlcNAc and LpxA are depicted as black dotted lines. Water molecules are represented as red spheres. Figure 5. UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc bound to P. aeruginosa LpxA active site. Monomers A and B of the P. aeruginosa complex are colored in cyan and green respectively, whereas the product is in yellow. A. The unbiased Fo-Fc map (contoured at 3.0 σ) unambiguously identifies the binding pose of UDP-3-O-(R-3hydroxydecanoyl)-GlcNAc in the active site. B. Details of the binding of the acyl chain in the hydrophobic cleft. C. Comparison of P. aeruginosa LpxA/UDP-3-O-(R-334 ACS Paragon Plus Environment

Page 35 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

hydroxydecanoyl)-GlcNAc complex to E. coli Lpxa/UDP-3-O-(R-3-hydroxymyristoyl)GlcNAc complex (light grey) showing the residues that orient the ligand in E. coli LpxA (E. coli ligand in light pink). His156 and Tyr158 of P. aeruginosa LpxA are also shown but not labelled. Residue labels are based on E. coli LpxA sequence. D. Comparison of P. aeruginosa LpxA/UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc complex to the L. interrogans LpxA/product complex (light grey) showing the residues that orient the ligand in L. interrogans LpxA, and also highlighting residues in chain B of P. aeruginosa LpxA that help explain differences in ligand pose between the two orthologs (L. interrogans ligand in light pink). Residue labels are based on L. interrogans LpxA sequence. Leu71, Lys72, His140 and Tyr158 of P. aeruginosa LpxA are also shown but not labeled. Figure 6. Acyl chain binding sites and hydrocarbon rulers. The two monomers at the dimer interface are colored in cyan and green respectively. Notice the hydrocarbon ruler is provided by a different monomer in P. aeruginosa LpxA in comparison to the other orthologs.

A. Surface model shows complementarity between the acyl chain and

hydrophobic pocket in the P. aeruginosa LpxA/UDP-3-O-(R-3-hydroxydecanoyl)GlcNAc complex structure, and supports the role of Met169 as the “hydrocarbon ruler”. B. L. interrogans LpxA complex shows Lys171 “hydrocarbon ruler”. C. E. coli LpxA complex shows His191 “hydrocarbon ruler”. Figure 7. Alternative backbone conformation for highly conserved residues Ala138 and Gly139. A. The P. aeruginosa LpxA/UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc (green/light green) complex superimposed to the E. coli/product complex structure (pink/light pink) shows opposing orientations for the protein backbone between highly 35 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 45

conserved residues Ala138/142 and Gly139/143, and the potential hydrogen bonds. B. The P. aeruginosa LpxA/UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc (green/light green) complex superimposed to the L. interrogans/product complex structure (purple/pink) shows opposing orientations for the protein backbone between highly conserved residues Ala138/137 and Gly139/138, and the potential hydrogen bonds C. Simulated-annealing omit map created by omiting residues 132-144 in the apo P. aeruginosa LpxA structure (1.81 Å) and contoured at 3.0 σ confirms that the backbone N of Gly139 is oriented away from the active site, while the backbone carbonyl of Ala138 is pointing towards the active site.


Page 37 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

LpxD*

LpxA* 3"OH"C10"ACP* 6*

5* O

HO 4* HO 3* 2* O

OH

OH

1* NH

O

UDP

O O HO

OH

O

HO

ACP*

3* O

O O

UDP

O

OH

O

HO

NH

Addi5onal* Enzymes*

3"OH"C12"ACP*

LpxC*

2* NH2 O

UDP

HO

O O HO

Lipid*A*

O

HO

ACP*

2* O

NH

O

UDP

HO

Figure 1. P. aeruginosa LpxA catalyzes the first and reversible step of lipid A biosynthesis.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 45

A*

900*

1800*

B* C3!!

C2!!

C1!!

L1*

C7!!

C5!!

C8!!

C4!!

C6!!

L2*

C*

D*

C9!!

L2*

C10!!

L1*

No*secondary*structure*assigned* β*helical*coil* Alpha*helix* Beta*strand*

Figure 2. P. aeruginosa LpxA apo crystal structure.


Page 39 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

A*

Asn194*

Arg201*

B*

Tyr158*

His140*

Ile148*

Phe166*

Glu196* His156*

Arg200*

Lys72* Leu71*

Gln157*

His121*

Lys72*

Tyr158*

Asp70*

Asp70* His140*

Leu71*

His121*

D*

C*

Arg205* Asn198*

Glu196*

Arg200*

Glu200* Arg201*

Met170*

Arg204*

Ile152*

Ile134* Leu75* Gln157*

His160*

Gln161*

Lys76*

Asp70* Asp74* Lys72*

Phe162* His144*

His125*

Figure 3. UDP-GlcNAc bound to P. aeruginosa LpxA active site. 39 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 45

Asp67*

Ac5ve*site* Gly48*

Lys47*

Ala30*

Pro24*

Asp11* Asp5* Arg7* Leu3*

Chain*B*

Chain*A* Secondary*site*

Figure 4. UDP-GlcNAc observed at a secondary binding site at the dimer interface. 40 ACS Paragon Plus Environment

Val22*

Page 41 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

A

B*

Asn194* Ile148* Glu196*

Phe166*

Arg200* Leu71*

Met169*

Asn133*

Tyr152*

Leu154*

C10*

Met169*

C10*

His156*

Val132*

Gly151*

Pro69*

Ala136*

Lys72*

Gln157* His118*

His118*

Ala138*

Asp70* His121*

Tyr158* His140*

Arg205*

C*

Asn198*

D

Asn193* Val195*

Glu200*

Met170*

Met165*

Arg204* Leu75* C14* Gln73*

His160* Gln161*

His122*

Lys76* Asp74* His125*

Phe162* His144*

Arg199* Phe147*

C12*

His155*

Leu70*

Phe157*

His117*

His139*

Gly71*

Gln68*

Gln156*

Asp69* His120*

Figure 5. UDP-3-O-(R-3-hydroxydecanoyl)-GlcNAc bound to P. aeruginosa LpxA active site.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 45

A* C10* Met169*

B* Lys171* C12*

C* C14* His191*

Figure 6. Acyl chain binding sites and hydrocarbon rulers. 42 ACS Paragon Plus Environment

Page 43 of 45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

A*

Gln73/Pro69*

His122/His118*

C*

C10*

Ala138*

C14* Ala138* Ala142* Gly139*

3.3*Å* Gly143* His125/His121*

B*

Gln68/Pro69*

His117/His118*

Gly139*

C10* C12*

Ala138* Ala137* Gly139*

3.3*Å* Gly138* His120/His121*

Figure 7. Alternative backbone conformation for highly conserved residues Ala138 and Gly139. 43 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 44 of 45

For Table of Contents Use Only

Structures of Pseudomonas aeruginosa LpxA reveal basis for substrate selectivity Emmanuel W. Smith,1 XiuJun Zhang,1 Cyrus Behzadi,1 Logan D. Andrews,2 Frederick Cohen,2 and Yu Chen*1 1

Department of Molecular Medicine, University of South Florida, 12901 Bruce B. Downs

Blvd., Tampa, FL 33612, USA 2

ACHAOGEN Inc., 7000 Shoreline Court, South San Francisco, CA 94080, USA

Met169*

His121*

*

Author to whom correspondence should be addressed: Yu Chen at (813) 974-7809; Fax (813) 974-7357; E-mail: [email protected]


Page 45 of 45

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

45

Structures of Pseudomonas aeruginosa LpxA Reveal the Basis for Its

Recommend Documents